Deep and accurate detection of m6A RNA modifications using miCLIP2 and m6Aboost machine learning

Area of Science:

• Molecular Biology
• Epigenetics
• RNA Biology

Background:

• N6-methyladenosine (m6A) is the most prevalent internal RNA modification in eukaryotes, regulating diverse RNA processing events.
• miCLIP is an antibody-based method for mapping m6A sites at single-nucleotide resolution, but suffers from challenges due to antibody cross-reactivity and false positives.
• Accurate m6A site identification is crucial for understanding its role in gene regulation.

Purpose of the Study:

• To significantly improve the accuracy and reliability of m6A site detection using an enhanced miCLIP technique (miCLIP2) combined with machine learning.
• To develop a robust computational pipeline to address false positives inherent in antibody-based m6A mapping.
• To create a universally applicable machine learning model (m6Aboost) for predicting genuine m6A sites.

Main Methods:

• Optimization of the miCLIP protocol (miCLIP2) to generate high-complexity libraries from reduced input material.
• Development of a computational pipeline calibrated using Mettl3 knockout cells to characterize m6A deposition patterns, including non-canonical sites.
• Training and application of a machine learning model, m6Aboost, utilizing experimental and RNA sequence features for m6A site prediction.

Main Results:

• miCLIP2 provides high-complexity libraries with reduced input requirements.
• The computational pipeline effectively mitigates false positives in antibody-based m6A detection.
• The m6Aboost model accurately predicts genuine m6A sites in miCLIP2 data, independent of DRACH motifs or Mettl3 depletion.
• Thousands of high-confidence m6A sites were identified across murine and human cell lines.

Conclusions:

• The combination of optimized miCLIP2 and the m6Aboost machine learning model represents a significant advancement in m6A identification.
• This methodology provides a robust and broadly applicable resource for studying m6A modifications and their functions.
• The findings facilitate deeper insights into the epitranscriptome and its regulatory roles.